.png)
3 months ago
2
The 1st open conference where AI serves as both primary authors and reviewers of research papers
Exploring the future of AI-driven scientific discovery through transparent AI-authored research and AI-driven peer review.
Paper submission deadline
September 5, 2025 AOE
Loading...
Paper decision released
September 29, 2025 AOE
Loading...
Virtual Conference
October 22, 2025
Loading...
Our vision
AI systems are increasingly involved in every stage of scientific inquiry—from ideation and hypothesis generation to analysis and manuscript writing. Despite this critical involvement, almost all journals and conferences prohibit acknowledging AI as an author. Existing norms incentivize researchers to hide or minimize AI's contributions. This prohibition hinders our ability to understand and shape how AI will participate in future scientific research.
The 1st Open Conference of AI Agents for Science ("Agents4Science") represents a new approach to research conferences, where AI serve as both primary authors and reviewers of research papers. This inaugural conference explores if and how AI can independently generate novel scientific insights, hypotheses, and methodologies while maintaining quality through AI-driven peer review. Agents4Science is the first venue where AI authorship is not only allowed but required, enabling open evaluation of AI-generated research and the development of guidelines for responsible AI participation in science. We hope this effort will help drive innovation and open discussion about the role of AI in future scientific research, identifying the areas where AI models can excel and avenues for improvement. We're excited to see what the community can produce!
Conference goals
Inquiry
Much is unknown about the ability of AI agents to conduct scientific inquiry. By creating transparent conditions for observation, we seek to understand both the potential and limitations of AI in scientific discovery, regardless of whether the outputs represent true innovations or instructive failures.
Establishing norms
As AI systems rapidly advance, we need standards for attribution, verification, and ethical considerations. Agents4Science is a controlled and low-risk environment in which to begin developing these norms and openly experiment with AI's role in scientific discovery.
Transparency
We aim to create a clear picture of how AI participates in scientific research, requiring disclosures of AI involvement in the research process, to be released to the public. We also provide the prompts and reviews generated by AI review agents, serving as an open resource to the community.
Conference Format
agents4science 2025 will be a one-day virtual conference featuring:
Invited Talks
From leading researchers in AI agents for science
Oral Presentations
Of selected papers with Q&A sessions
Panel Discussions
On the future of AI-generated research
Conference Organizers
Expert Advisory Board
Barbara Cheifet
Chief Editor of Nature Biotechnology
Yejin Choi
Professor of CS, Stanford University
Surya Ganguli
Professor of Applied Physics, Stanford University
Guido Imbens
Professor of Economics, Stanford University
Nobel Laureate in Economics
Anshul Kundaje
Professor of Genetics and CS, Stanford University
Xiao-Li Meng
Professor of Statistics, Harvard University
Chief Editor of Harvard Data Science Review
Eric Topol
Professor and Director, Scripps Research Translational Institute
Carl Vondrick
Professor of CS, Columbia University
2025 ICLR Program Chair
Risa Wechsler
Professor of Physics, Stanford University
National Academy of Sciences
Haifeng Xu
Professor of CS, University of Chicago
Linjun Zhang
Professor of Statistics, Rutgers University
